Research Quality and Validity

Creswell (2008), suggests that researchers need to convey the steps taken in their study to check for the accuracy and credibility of their finding. The exploratory case study strategy has been used which is supported by the constructivist paradigm as a qualitative research method.

Qualitative researching engages researchers to explore a wide array of dimensions of the social world, including the texture and weave of everyday life, the understandings, experiences and imaginings of research participants, the ways that social processes, institutions, discourses or

56 relationships work, and the significance of the meanings that they generate (Mason, 2002).

This is done using in-depth, refined, context based, multidimensional and complex methodologies that intimately connect context with explanation of how things work. Therefore, it is capable of producing well founded cross-contextual generalities (Ibid). However, qualitative research has its own challenges. Some scholars argue that qualitative research is of poor standard and has no clearly defined set of quality criteria available for judging it hence the quality is uncertain (Hammersey, 2007). This implies that it is potentially an assembly of anecdote and personal impressions, strongly subject to researcher bias, lacks reproducibility and generalisability. Therefore, a good qualitative research design need to provide reliability, validity and generalisability of the construct (Ibid).

3.5.1 Validity

Qualitative validity involves employing certain procedures in research to check for accuracy of findings from the stance of the researcher, participants and readers of the research account (Creswell & Miller, 2000). In other words, it shows the trustworthiness or creditability of a research. It means your research is valid, if you are observing, identifying or measuring what you say you are. For example, in researching on cross cultural conflicts in business relationships, I will need to be able to show that my data on analysis of ideas, norms and negotiations about conflict did relate to the concepts. Validity is often associated with the

‘operationalization’ of concepts and the need to be able to demonstrate that the concepts can be identified, observed or ‘measured’ in the way claimed (Mason, 2002). Even though some scholars argue that the subjectivity of qualitative research makes it harder to accurately show validity, incorporating validity strategies in qualitative research is a strength (Shields &

Rangarajan, 2013; Creswell, 2009). It is therefore important to work out how well a particular method and data source might illuminate the concepts in the research. Guba and Lincoln (1981), Krefting (1991) and Creswell (1998) suggest that the trustworthiness or validity of qualitative research can be established by using four strategies: credibility, transferability, dependability and conformability. These are constructed parallel to the analogous quantitative criteria of internal and external validity, reliability and neutrality.

3.5.2 Credibility

Credibility in qualitative research can be defined as the extent of believability and trustworthiness of data collected and data analysis process. It is analogous to internal Validity in quantitative research which is the approximate truth about inferences regarding cause-effect or causal relationships (Yin, 2014). It means is that you have evidence that what

57 you did in the study (i.e., the program) caused what you observed (i.e., the outcome).

Credibility exists when the research findings reflect the perceptions of the people under study Credibility is important in qualitative research, as researchers are able to demonstrate the reality of the participants through detailed description of the discussion. Most rationalists would propose that there is not a single reality to be discovered, but that each individual construct a personal reality (Smith & Ragan, 2005). To increase credibility, the following strategies were implemented;

1) “Rich” data: Becker (1970) argues that, both long-term involvement and intensive interviews enables a researcher to collect “rich” data, that is, well detailed and varied data that provide a full and revealing picture of situation. In interview studies, such data generally require verbatim transcripts of the interviews (Maxwell, 2005). Rich data was collected by detailed notes taking during the interview. Whilst, the interviewer was interviewing the informant, the other two members of the team were taking very detailed notes and not just what they felt was significant. After each interview, notes were shared, compared and discussed to ensure information recorded was similar, detailed and accurate. I believe this process enabled collection of rich data and hence increases the credibility of the data.

2) Respondent validation: Respondent validation can be referred to as “member” checks (Bryman, 1988). It requires systematically soliciting feedback about one’s data and conclusions from the participants (Maxwell, 2005). It is used to determine the accuracy of the qualitative findings by taking the final report or specific descriptions or themes back to participants and determining their perception about the accuracy (Creswell, 2008). This helps to rule out the possibility of misinterpreting the meaning of what participants say and do and identifies any biases and misunderstandings of what was observed (Maxwell, 2005). As mentioned above, proper field notes were sent to each participant to review, to make sure, there were no misunderstandings or misinterpretation about their perceptions of the concepts. There was no negative feedback from the participants. Even though, participants’ feedback is no more inherently valid than their interview responses, I consider it as evidence regarding the validity of my account (Hammersley & Atkinson, 1995).

3) Triangulation of data: It involves collecting information from a diverse range of individuals and settings, using different methods (e.g., observation, interview, document analysis, etc.) (Maxwell, 2005), in order to gather multiple perspectives on the same issue so as to gain a more complete understanding of the phenomena. This

58 strategy provides a better assessment of the generality of developed explanations and reduces the risk of chance associations and of systematic biases as a result of using a single method (Ibid). It is one of the most important ways to improve the trustworthiness of qualitative research findings and improve confidence. Data was collected through sources such as interviews and document analysis. In-depth interviews were conducted which began with open ended questions and progressed to more structured questions. This enabled the participants to comfortably reveal their perception about the phenomenon. Other methods used were, member checking, clarification of biases, and document analysis. This enabled the use of multiple referents to draw conclusions. Triangulating the outcome of the interviews with document analysis enhances the credibility of this research.

The above procedures were the easier and the most-cost effective ways I could use to enhance the credibility of my research.

3.5.2 Generalizability

Generalizability involves the extent to which a wider claim can be made on the foundation of a research and analysis study (Mason, 2002). The generalizability of qualitative studies is usually based on the development of a theory that can be extended to other cases (Becker, 1991;

Ragin, 1987). It deals with the problem of knowing whether a study's findings are generalizable beyond the immediate case study. Yin (1994) refers to this as analytical, generalization where, the investigator is striving to generalize a particular set of results to some broader theory. The generalization occurs when qualitative researchers study additional cases and generalize findings to the new cases (Creswell, 2008). In other words, research findings can only be generalisable if they fit into new contexts outside the actual study context. The subjectivity of the researcher as a key instrument of qualitative research makes generalisability difficult and threatens the validity of inferences from research data. However, Seale (1999) proposes that generalisability can achieved by providing a detailed, rich description of the settings studied to provide the reader with sufficient information to be able to judge the applicability of the findings to other settings that they know. Yin (2003) further adds that good documentation of qualitative procedures, such as a protocol for documenting the problem in detail and the development of a thorough case study database can achieve generalisability of qualitative case study research.

This study uses the cross-case approach and therefore will make use of inferential generalisation which is generalisation in terms of nature and diversity of phenomena (Ritchie

59

& Lewis, 2003). Therefore, proper documentation and justification of methodological approach, detailed description of critical processes and procedures used to mould and associate meaning to phenomena is required. Using a qualitative methodological approach makes it challenging to generalise. Also, the informants or companies interviewed may not actually represent the Norwegian-Brazilian business relationship. However, I have provided a rich and detailed description of the study so that anyone interested in generalisation will have a solid framework for comparison (Merriam, 1988) and make their own judgement regarding the transferability of the research outcome. Therefore, the generalizability of this outcome has to be resolved by the reader based on the closeness of his or her context to that of the researcher. 3.5.3 Reliability

Merriam (1998) defines reliability as to the extent to which research findings can be replicated.

Qualitative reliability shows uniformity of a researcher’s approach across different researchers and projects (Gibbs, 2007). Yin (2003) suggests that qualitative researchers can check if their approaches are consistent and reliable by documenting as many of the steps of their procedures as possible and setting up detailed case study protocol and data base. Similarly, Gibbs (2007) suggests that reliability procedures include checking transcripts for transcription mistakes, ensuring consistency in definition of codes and comparing independently derived results.I have tried to provide detailed description of my procedures and choices during the research process in order to increase reliability and make it possible for other researchers to evaluate my study.

However, Thomas (2010) argues that reliability is problematic and is practically impossible as human behaviour is dynamic, highly contextual and changes continuously based on various influencing factors. Furthermore, possibility of different perceptions of the phenomenon by the participants in a research study implies that a similar study with different informants may not necessary produce the same findings. The quality of inferences is subjective making reliability pointless in qualitative study. This research was conducted in a similar manner to previous researcher (Granli, 2012). His results are similar to mine which helps to increase reliability.

3.5.4 Confirmability

Confirmability can be defined as neutrality or objectivity of data (Polit et al., 2001). It shows the degree to which the research findings can be confirmed by others. Seale (1999) argues that auditing could also be used to establish confirmability in which the researcher makes the provision of a methodological self-critical account of how the research was done. This involves tracing the researcher’s path of arriving at the constructs, themes and interpretation hence

60 enhancing confirmability. Also auditing can be enabled by the researcher archiving collected data in a retrievable form which can be accessible to other researchers in case his/her findings are challenged. All the data collected for this research have been well archived, hence increasing the confirmability.

3.5.5 Reflexivity

In qualitative research the researcher is both the researcher and the participant and can therefore not be divorced from the phenomenon under study. According to Parahoo (1997), reflexivity is a continuous process whereby researchers reflect on their preconceived values and those of the participants. Holloway and Wheeler (2002) add that researchers should reflect on their own actions, feelings and conflicts experienced during research. To achieve credibility of the study, the researcher adopted a self-critical stance to the study, the participants, their role, relationships and assumptions. Reflexivity is not easy because identifying one’s own preconception is difficult. Reflexibility helps the researcher with monitoring, self-injunction, and reduces the risk of bias. In this study, I wrote down any feelings, preconceptions, conflicts and assumptions I had about the study. I tried to prevent my cultural perceptions from influencing my interpretation of the data. I took steps to be objective and not misinterpret data. This enabled self-monitoring to prevent bias and increase objectivity.